The effects of ageing on self-reported aggression

Anuncio
Psicothema 2015, Vol. 27, No. 3, 209-215
doi: 10.7334/psicothema2015.32
ISSN 0214 - 9915 CODEN PSOTEG
Copyright © 2015 Psicothema
www.psicothema.com
The effects of ageing on self-reported aggression measures are partly
explained by response bias
Andreu Vigil-Colet, Urbano Lorenzo-Seva and Fabia Morales-Vives
Universitat Rovira i Virgili
Abstract
Background: Recent studies have suggested that the age-personality
relationship may be partly explained by age-related changes in
response bias. In the present study, we analysed how age affected social
desirability and acquiescence, and how this effect impacted the ageaggression relationship. Method: We used the Indirect-Direct Aggression
Questionnaire, which provides response bias and physical, verbal and
indirect aggression scores independently of each other. We applied this
test to a sample of 616 individuals aged between 18 and 96 (M = 49.24,
SD = 24.81) and analysed the relationships between age and aggression
measures with and without response bias. Results: We found that social
desirability and acquiescence increased by between one and two standard
deviations between adulthood and old age. This affected the age-aggression
relationship for all aggression scales and, especially for verbal and indirect
aggression, whose relationships with age decreased from r = -.192 and r
= -.309 to r = .012 and r = -.159, respectively, when response biases were
controlled. Conclusions: When response bias and, in particular social
desirability, are not controlled, elderly people tend to show aggression
scores that are considerably lower than their true aggression levels.
Keywords: Aggression, response bias, personality.
Resumen
Los efectos del envejecimiento en las medidas de agresividad se explican
en parte por los sesgos de respuesta. Antecedentes: estudios recientes
han sugerido que las relaciones entre la edad y la personalidad pueden
deberse parcialmente a los sesgos de respuesta. En el presente estudio
hemos analizado los efectos de la edad en los sesgos de respuesta y
cómo estos cambios afectan las relaciones entre la edad y la agresividad.
Método: aplicamos el Cuestionario de Agresividad Indirecta-Directa que
proporciona puntuaciones libres de sesgo en agresividad física, verbal e
indirecta a 616 individuos con edades comprendidas entre los 18 y los 96
años (M= 49.24; SD= 24.81), analizando las relaciones entre las medidas de
agresividad y la edad en presencia o ausencia de los efectos de los sesgos de
respuesta. Resultados: ambos sesgos de respuesta presentaron incrementos
de entre una y dos desviaciones típicas a lo largo del ciclo vital, afectando
a la relación entre edad y todas las escalas de agresividad, especialmente
en el caso de la agresividad verbal y la indirecta, cuyas relaciones con
la edad disminuyeron desde r= -.192 y r= -.309 hasta r= .012 y r= -.159,
respectivamente, al controlar los sesgos de respuesta. Conclusiones:
cuando no se controlan los sesgos de respuesta y, especialmente, la
deseabilidad social, los ancianos tienden a mostrar niveles de agresividad
muy inferiores a sus niveles reales.
Palabras clave: agresión, sesgos de respuesta, personalidad.
Changes in personality in the course of ageing have often been
considered a consequence of better emotional regulation and increased
maturity, leading to higher scores in personality dimensions such as
agreeableness or emotional stability in older people (Caspi, Roberts,
& Shiner, 2005; Soto, John, Gosling, & Potter, 2011; Terracciano,
McCrae, Brant, & Costa, 2005). Referring to these changes in older
people, Marsh, Nagengast & Morin (2013) proposed the term “dolce
vita effect”, arguing that old people are happier, more self-contented
and self-centred because they are less worried about their career,
work, etc., and therefore have a more positive attitude towards life.
Furthermore, various studies have shown that the ageing
process also has positive effects on aggressive behaviour, with age
Received: January 28, 2015 • Accepted: April 28, 2015
Corresponding author: Andreu Vigil-Colet
Research Center for Behavior Assessment
Universitat Rovira i Virgili
43007 Tarragona (Spain)
e-mail: [email protected]
being negatively correlated with aggression, more specifically with
physical aggression, although a negative relationship has also been
found for verbal aggression and other aggressive behaviours such
as aggressive driving (Archer, 2004; Gerevich, Bácskai, & Czobor,
2007; Krahé & Fenske, 2002; O’Connor, Archer, & Wu, 2001).
These negative relationships have been explained by authors
such as Walker and Richardson (1998) and Walker, Richardson
and Green (2000) taking into account that the potential danger
associated with direct aggression might discourage the use of
such approaches in older people, leading them to choose less risky
strategies based upon indirect aggression through the manipulation
of the social contacts within their interaction network.
For authors such as Marsh et al. (2013), the ageing process in
older people leads to higher levels of agreeableness and emotional
stability and lower levels of openness and extraversion. These
studies have mainly been carried out using self-reports, which are
sensitive to response biases such as social desirability (SD) and
acquiescence (AC). These response biases are positively related to
age, so various authors have recently suggested that age-personality
209
Andreu Vigil-Colet, Urbano Lorenzo-Seva and Fabia Morales-Vives
relationships may be reflecting not only changes in personality but
also the effect on test scores of changes in response bias related to
age, and especially to changes in SD (Soubelet & Salthouse, 2011;
Vigil-Colet, Morales-Vives, & Lorenzo-Seva, 2013).
The effects of age-related changes in social desirability have
often been reported, and it seems that there is a tendency among
older people to present themselves favourably, showing higher
scores in SD than young and middle-aged adults (Dijkstra, Smit, &
Comijs, 2001; Ray, 1988; Stöber, 2001). These effects also seem
to be of a considerable magnitude. Stöber (2001), for instance,
reported differences between older people and adolescents of 1.75
standard deviations. A similar effect of 1.9 standard deviations was
reported by Vigil-Colet, Morales-Vives and Lorenzo-Seva, (2013)
in a sample aged 13 to 95. Bearing in mind the magnitude of these
increases, it is possible that the lifetime evolution of personality
scores for traits that are more affected by SD may be partly
reflecting the effect of these increases in. Furthermore Vigil-Colet,
Morales-Vives and Lorenzo-Seva (2013) found that the increases
in response bias associated with age were slightly greater for
women. Consequently, these increases also had a larger effect on
the age/personality relationship.
In this respect, Soubelet and Salthouse (2011) analyzed the
relationships between the personality traits of the five-factor model
and age, finding that age was negatively related to neuroticism and
positively related to consciousness and agreeableness. However,
these relationships were reduced when SD was controlled and, in
the case of agreeableness, the relationship actually disappeared. As
a result, they concluded that some of these age-related differences
in personality and other variables, such as affect measures, could
be attributable to age-related increases in SD.
In the study by Soubelet and Salthouse (2011), SD effects were
controlled using the traditional procedure whereby the measures
of interest and an SD scale are administered together, with the SD
scale then being used to partial out the SD effects.
This procedure has two main limitations that should be taken into
account. Firstly, it involves administering a second scale together
with the scale of interest and, secondly, it assumes that the items
on the SD scale and the content scale are parallel measurements,
which is highly improbable (Leite & Cooper, 2010).
A different approach which may overcome these limitations
was proposed by Ferrando, Lorenzo-Seva and Chico (2009), based
on a procedure that can be used to control not only SD but also AC
and that provides bias-free scores for the content variables. To this
end the method first identifies a factor related to SD by using a few
items that are taken as SD markers. The inter-marker correlation
matrix is then analysed using factor analysis and the corresponding
loading values are used to compute the loading values of the content
items for the SD factor. The variance explained by the SD factor
is then removed, leaving a residual inter-item correlation matrix
without SD effects. This matrix is then analysed and the variance
due to acquiescent responding is removed from the content items.
This analysis can be made using a method proposed by LorenzoSeva and Ferrando (2009) which identifies acquiescence response
in a style factor in partially balanced scales (i.e., where the number
of content reversed items does not equal the number of directly
worded items). The procedure analyzes the inter-item correlation
matrix and identifies a common factor related to acquiescence. The
variance caused by acquiescent responding is then removed from
the content items to obtain a second residual inter-item correlation
matrix. This second residual inter-item correlation matrix is free
210
of both SD and AC. Finally, a classical exploratory factor analysis
is performed on this residual matrix to determine item loadings
on content factors. This allows individuals’ factor scores to be
obtained free of response bias effects. It should be noted that these
SD and AC factors are orthogonal to each other and to the content
factors; hence the removal of SD and AC bias does not affect the
valid variance of the content factors.
Vigil-Colet, Morales-Vives, Camps, Tous and Lorenzo-Seva
(2013) used this method to develop a test to assess personality within
the framework of the five-factor model. This test, known as the
Overall Personality Assessment Scale (OPERAS), provides bias-free
scores for the five personality scales plus SD and AC scores for each
individual. It is also possible to compute scores for each individual
without removing bias effects, thereby enabling the age-personality
relationship to be analysed using personality scores with and
without response bias effects. When the age-personality relationship
was analysed for both types of score, Vigil-Colet, Morales-Vives
and Lorenzo-Seva (2013) found that response bias affects the
relationships of conscientiousness (CO) and agreeableness (AG)
with age, with these dimensions being the ones most affected by SD.
In the case of AG, when the response biases were not controlled,
the relationship with age was positive, but when the response bias
effects were removed, the relationship became negative.
Bearing in mind that response bias seems to affect the personalityage relationship and that this effect is especially relevant for those
dimensions most affected by SD, the question is whether response
bias and in particular SD may also be affecting the relationships
of age with other personality measures that are highly impacted
by SD. This is especially relevant in the case of self-reported
aggressiveness, because aggressiveness is related to age, and selfreported measures of aggressiveness are highly affected by SD.
Although it is possible that age may affect aggression strategies
in the sense suggested by Walker and Richardson (1998), it
also needs to be taken into account that if aggressive measures
are highly impacted by SD, then changes in SD associated with
age may also partly explain the relationships between age and
aggression measures. In this respect, it is also important to take
into account that most of the studies that analyse the relationships
between age and aggression have been conducted in samples aged
under 60, and the few studies carried out with older samples up
to age 95 have reported that quite high levels of impulsivity and
aggression may also be found in some older individuals (MoralesVives & Vigil-Colet, 2010; 2012).
Several studies have shown that there is a substantial relationship
between SD and aggression measures. Biaggio (1980) and Selby
(1984), for example, reported correlations between the BussDurkee Hostility Inventory (BDHI) and the Marlow-Crowne Social
Desirability (MCSD) in the range of r=-.3 to -.5. Similar results
have been found for the Buss and Perry Aggression Questionnaire
(BPAQ) (Becker, 2007; Harris, 1997; Morren & Meesters, 2002),
and for indirect aggression measures (Selby, 1984).
All the above results regarding the relationships between SD and
aggression measures were obtained by analysing the relationships
between the scores of self-reported aggression and SD scales. A
different approach was adopted by Vigil-Colet, Ruiz-Pamies,
Anguiano-Carrasco and Lorenzo-Seva (2012), who administered
two aggression measures, the BPAQ and a Spanish adaptation of
the Indirect Aggression Scale (Anguiano-Carrasco & Vigil-Colet,
2011) together with four SD items which, following the procedure
developed by Ferrando et al. (2009), made it possible to obtain
The effects of ageing on self-reported aggression measures are partly explained by response bias
the loadings on SD of the items of those aggression measures.
The results obtained by Vigil-Colet et al. (2012) showed that both
direct and indirect aggression measures were highly impacted by
SD, with physical and verbal aggression scales for more than half
the items at λ >.30 on the SD factor, while indirect aggression
showed a lower but still relevant impact of SD, with 30% of items
at λ >.30.
Taking into account the relationships between age, aggression and
SD set out above, the purpose of this study is to analyse the impact
of response bias on the relationship between age and self-reported
aggression and, more specifically, whether SD may partly explain the
observed relationships between aggression and age. To this end, we
will use a new aggression measure, the Indirect-Direct Aggression
Questionnaire recently developed by Ruiz-Pamies, Lorenzo-Seva,
Morales-Vives, Cosi and Vigil-Colet (2014), following the method
developed by Ferrando et al. (2009). This test will enable us to
obtain measures of physical, verbal and indirect aggression free of
SD and AC, together with scores on both response biases. It will
also be possible to compute scores for the aggression scales without
removing response biases, thereby enabling their impact on the
relationships between age and aggression to be evaluated. We will
also test whether these effects have any sex-related differences, as
they do in overall personality measures.
The main hypothesis of the study is that if self-reported
aggression measures are impacted by response biases, and these
biases increase with age, the negative relationship between age and
aggression should decrease when response biases are controlled,
especially for those aggressive behaviours that are less related
to potential physical danger to older people, such as verbal and
indirect aggression.
Method
Participants
The sample consisted of 616 adults ranging between 18 and 96
years of age (M= 49.24; SD= 24.81). 228 individuals were aged
between 18 and 37, 141 between 38 and 57, 126 between 58 and
77, and 118 between 78 and 97. 60.2% of the sample was female.
Taking into account that we had to tap a wide age range, various
strategies were used to collect the sample, because it is difficult
to find one specific population comprising people from late
adolescence to old age. To this end, the test was administered to: (1)
university students in their classrooms (52% females); (2) workers
from different companies during training activities; (55% females)
(3) participants in academic university activities for graduates/
adults (64% females); and (4) older people in nursing homes and
on specially-designed university courses (67% females).
Instruments
Aggression measures
The Indirect-Direct Aggression Questionnaire (I-DAQ; RuizPamies et al., 2014). This test gives SD and AC-free scores for the
factors Physical Aggression (PA), Verbal Aggression (VA), Indirect
Aggression (IA) and Overall Aggression (OA), plus SD and AC
scores for each individual. The questionnaire consisted of 6 items
measuring PA, 7 items measuring VA and 10 items measuring IA
plus 4 items used as SD markers. The factors measured by I-DAQ
have appropriate factorial (rθθ =.83; rθθ =.77 and rθθ =.78) and testretest (rtt =.71; rtt =.61 and rtt =.69) reliabilities for PA, VA and
IA, respectively. The test also has good convergent and criterion
validity with other aggression and personality measures. The test
gives the scores for each content scale and for response biases on
the T-scale. Detailed information about the items and psychometric
properties of the test can be found at http://psico.fcep.urv.cat/tests/
idaq/en/index.html.
Cognitive measures
The Mini-Mental State Examination (MMSE) (Folstein,
Folstein, & McHugh, 1975). We administered the Spanish
version of the MMSE (Lobo et al., 1999), which is a quick test
for screening cognitive function deficits in older people and one
of the most widely-used questionnaires for this purpose. The test
consists of 23 items and their scores range from 0 to 35 points. We
administered the aggression questionnaire only to individuals with
scores above 23 in the MMSE, which is the recommended cut-off
point for detecting people at risk of having dementia. This cut-off
point gives good sensitivity (90%) and specificity (75%), with an
area under the ROC curve of .92. Twenty-four individuals were
not included in the study because they had scores under the cut-off
point used to detect possible cases of dementia.
Procedure
The test was administered by a psychologist to students and
workers in groups of between 20 and 40 individuals. In the case
of older people, the test was administered individually or in small
groups (up to five people) by a psychologist experienced in dealing
with older people. These participants were also tested to exclude
people with dementias. The aggression questionnaire was selfadministered and the participants were volunteers and were asked
for no information that could have identified them, and their
anonymity was therefore guaranteed. They were not paid for their
participation.
Data analysis
We computed the factor scores for each participant on the
aggression measures for the factor solution obtained before and
after SD and AC effects had been removed. This procedure makes
it possible to compare the relationships between age and aggression
measures when aggression scores are free of or affected by response
bias. It also provides SD and AC scores for each individual. With
these scores, we computed the product moment correlations between
age and response bias. We also computed the correlations between
aggression measures and age for both corrected and uncorrected
scores. To graphically represent the relationship between these
variables, we split the individuals into four age groups because
these relationships and the effects of removing response biases are
displayed more clearly than with a scatter plot.
Results
Table 1 shows descriptive statistics for response bias and content
scores with and without response bias effect, which allow us to
evaluate the impact of response bias on the relationships between
age and self-reported aggression measures.
211
Andreu Vigil-Colet, Urbano Lorenzo-Seva and Fabia Morales-Vives
Table 1
Descriptive statistics for bias and content scales across four age groups with and without response bias (Scores in T-Scale)
Without response bias
Social
desirability
Acquiescence
Physical
Verbal
With response bias
Indirect
Overall
Physical
Verbal
Indirect
Overall
Age
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
18-37
47.0
08.2
49.9
10.3
52.0
10.0
48.9
09.7
51.3
10.0
51.2
9.5
52.1
10.1
50.2
09.3
52.4
10.2
52.1
09.3
38-57
53.2
10.2
49.9
09.7
47.7
08.7
48.5
10.2
49.5
09.2
48.2
9.1
47.8
08.8
47.2
10.4
48.1
08.9
47.1
08.9
58-77
60.9
08.6
56.2
10.5
48.3
08.7
51.6
12.5
47.4
09.6
48.3
9.8
44.9
09.4
48.5
12.5
46.1
09.2
45.4
10.3
78-97
64.8
06.6
55.8
08.4
46.4
07.5
48.6
12.1
47.7
10.2
46.6
9.7
42.7
08.5
44.3
11.4
44.8
08.6
42.2
09.0
It can be seen that the response bias increased with age, and that
this increase was greater for SD than for AC, with SD increasing
by almost two standard deviations from the youngest to the oldest
group (see Figure 1). As far as aggression measures are concerned,
physical, indirect and overall aggression decreased with age, but
this decrease was much greater when the response biases were
not controlled, reflecting a possible effect of response bias on
aggression scores across age groups. This effect was especially
relevant for verbal aggression, which apparently showed no age
effect when the response biases were controlled but did show an
age effect when they were not.
Table 2 shows the correlations between age, response bias and
aggression measures. Age is positively related to both response
biases, its relationship being much higher for SD than for AC.
Indeed age accounts for more than 40% of the variance of SD.
In the case of aggression measures, when response biases are not
controlled, age is negatively related to all aggression measures.
However, when we computed the correlations using scores free of
response bias, all these relationships decreased and in the case of
verbal aggression disappeared. Figure 2 shows these effects across
age groups. It can be seen that the age effects on aggression using
corrected and uncorrected scores are quite similar for the first two
age groups, but in the two older groups, the uncorrected scores
show considerably lower aggression levels than the corrected
scores.
70,00
In order to determine which response bias was affecting the
age-aggression relationship, we computed partial correlations
between age and aggression scales controlling for SD and AQ.
Table 2 shows these correlations, and it can be seen that when
the AC effects were partialled out, the size of the correlations
increased. However, when the SD effects were controlled, the size
of the correlation between age and physical aggression decreased,
disappearing for the other measures, thereby indicating that the
main cause of the differences found when age was related with
corrected and uncorrected aggression scores was SD.
Taking into account the fact that some studies have proposed
differences in levels of SD associated with gender, we analysed the
age-aggression relationships for men and women separately. The
results are shown in Table 3. It can be seen that the age-aggression
relationship was quite similar for both men and women, and
that the differences between the observed correlations were only
significant for physical and verbal aggression, but with a low effect
size. The only relevant effect was that AC was more related to age
in women than in men, while the much larger relationship of SD
with age was very similar for both sexes.
Discussion
The results set out above confirm the hypothesis that response
biases and, more specifically SD, partly explain the relationships
between age and aggression. Furthermore, these results are
similar to those of other studies that showed that SD and AC were
Table 2
Correlations of age with response bias and content scales, with and without
response bias effects on content scores and controlling for acquiescence and
social desirability
65,00
60,00
Acquiescence
.652**
.270**
Age
55,00
50,00
Age
45,00
18-37
38-57
Social desirability
58-77
78-97
Scores controlling bias
*
Physical
Verbal
Indirect
Overall
-.220**
.012
-.159**
-.180**
**
Scores with response bias
-.373
-.192
-.309
-.383**
Controlling aquiescence
-.286**
-.240**
-.420**
-.402**
Controlling social
desirability
-.279**
.067
.002
-.088*
Acquiescence
Figure 1. Social desirability and acquiescence across age categories
212
Social desirability
p<.05; ** p<.01
**
**
The effects of ageing on self-reported aggression measures are partly explained by response bias
53,00
55,00
51,00
53,00
51,00
49,00
49,00
47,00
47,00
45,00
45,00
43,00
43,00
41,00
41,00
18-37
38-57
58-77
18-37
78-97
38-57
58-77
Physical
Indirect
Physical with bias
Indirect with bias
55,00
55,00
53,00
53,00
51,00
51,00
49,00
49,00
47,00
47,00
45,00
45,00
43,00
43,00
41,00
78-97
41,00
18-37
38-57
58-77
78-97
18-37
38-57
58-77
Overall
Verbal
Overall with bias
Verbal with bias
78-97
Figure 2. Evolution of physical, verbal, indirect and overall aggressiveness across age categories for bias-corrected and bias-uncorrected scores
Table 3
Comparisons for the correlations of age with response bias and content scales
for men and women using Fisher’s r-to-z transformation test
Variable
Controlling bias
With response bias
*
p<.05; ** p<.01
Men
Women
Social desirability
.596**
.682**
n.s.
Acquiescence
.115
.387**
<.01
Physical aggression
-.287**
-.124*
<.05
Verbal aggression
-.163*
-.087
<.05
Indirect aggression
-.218**
-.077
n.s.
Overall aggression
-.196**
-.131
n.s.
**
**
p (z test)
Physical aggression
-.363
-.357
Verbal aggression
-.064
-.265**
<.05
n.s.
Indirect aggression
-.388**
-.214**
<.05
Overall aggression
-.378**
-.36**
n.s.
affected by age, especially in older people (Becker, 2007; Harris,
1997; Morren & Meesters, 2002). In this respect, SD is the most
affected response bias, showing an increase of nearly two standard
deviations in the course of adulthood. However, our main objective
was not to establish the magnitude of these increases but to analyse
the effects of the increases in response bias on the age-aggression
relationship. In this respect, our results seem to show that response
bias accounts for a relevant part of the variance between age and
the scores from aggression self-reports. These results are along the
same lines as the ones reported for the effects of response bias
on the age-personality relationships within the framework of the
Big Five model of personality (Soubelet & Salthouse, 2011; VigilColet, Morales-Vives, & Lorenzo-Seva, 2013).
The results obtained by comparing the relationships with age of
corrected and uncorrected scores on the IDAQ scales showed that
when response biases were controlled, the negative relationship
between age and aggression is much smaller than the relationship
found when they are not controlled. This effect is almost the same
213
Andreu Vigil-Colet, Urbano Lorenzo-Seva and Fabia Morales-Vives
for both sexes and is especially relevant in the case of verbal
aggression, which, using bias-free scores, apparently remains
stable throughout adulthood, although it is negatively related to
age when the response bias effects are not controlled.
One interesting question is to determine which response bias
is responsible for the decrease observed in the age-aggression
relationships. When the effects of SD and AC were individually
controlled, it was observed that the correlations decreased in the
case of SD but increased in the case of AQ. These results, which
may seem contradictory, could be explained by taking into account
that response bias may be considered content variance or error
variance.
This last issue is a long-standing subject of discussion in response
bias research. There are two main points of view regarding SD
and AC. The first considers them as measurement error or artefacts
that are of no substantive interest, while the second considers
them as individual difference variables that have a certain degree
of stability and consistency (Ferrando et al., 2009). These points
of view have different predictions about how the removal of
biases from the content variable affects its validity coefficients.
If response bias is measurement error, then when it is removed,
the coefficients should increase, whereas if response bias is a
meaningful trait, then when it is removed, the correlations should
decrease. As Paulhus (1991) has pointed out, there are various SD
measures with an appropriate reliability and temporal stability,
but the few AC measures that exist do not have the appropriate
psychometric properties. Furthermore, different studies have found
relationships between SD and personality measures, which is not
compatible with the idea of SD as measurement error, but these
kinds of relationship have not been found for AC (Li & Bagger,
2006; McCrae & Costa, 1983; Ones, Viswesvaran, & Reiss, 1996).
Consequently it seems that SD is a meaningful trait whereas AC
is more related to error variance. Nevertheless, the present study
has not been specifically designed to test this hypothesis, so further
research is needed so that it can be verified and generalized to other
personality domains.
These differences between SD and AC may explain the observed
effects on the age-aggression relationship when response biases are
controlled. In the case of AC, when response bias is removed, the
correlation between age and aggression measures tends to increase
because the associated decrease in measurement error in aggression
measures results in an increase in correlation coefficients. However,
SD seems to be a meaningful trait that affects aggression measures
and increases with age because, when response bias is controlled,
the age-aggression relationship decreases.
Nevertheless, although age-aggression relationships are affected
by response biases, the fact is that even when response biases
are controlled, there is a tendency to find a negative relationship
between age and aggression measures, though this is smaller than
that reported when response biases are not controlled. When this
relationship is analysed for each kind of aggression, we can see that
physical aggression clearly decreases from young adulthood until
age 50-60 and that this decrease is lower for indirect aggression
and non-existent in the case of verbal aggression. These results
are similar to those reported by Walker and Richardson (1998) and
Walker et al. (2000), who suggested that older people use strategies
to minimize the effect-danger ratio, which, taking into account the
decline in strength and health of older adults, may favour the use of
aggressive behaviours that involve no risk to health, unlike in the case
of physical aggression. More specifically, Walker and Richardson
(1998) suggested that these processes should be reflected in a
decrease in the likelihood of direct forms of aggression (especially
physical aggression) compared to other forms of aggression
(especially indirect aggression). The results reported above follow
similar lines, showing that physical aggression decreases in favour
of more indirect aggression as individuals get older. However, our
results showed that verbal aggression remains stable throughout
the life span, so the differences in direct aggression associated with
age may be due mainly to variations in physical aggression levels.
The results set out above have relevant implications for
assessment in older people and especially for the assessment of
aggressive behaviour. Our study, like previous studies, shows that
the highest levels of SD and AC are reached in older people, and
therefore the results of self-reports that do not incorporate any
correction for response bias should be viewed with caution. Older
people may show themselves to be less aggressive than they really
are, making it difficult to carry out a valid assessment of their
aggression levels. The main problem arises from the fact that old
people have the highest levels of SD and AC, and their response bias
scores also show considerable variability. Therefore, if variance
due to response bias is not removed, any comparison of individuals
with their reference group may be contaminated by these sources of
error, and clinicians may take wrong decisions. This is especially
relevant bearing in mind that there are a considerable number
of individual differences in faking aggression questionnaires,
especially for indirect and verbal aggression (Anguiano-Carrasco,
Vigil-Colet, & Ferrando, 2013).
The present study suffers the classical problem of cross-sectional
studies and the possible effects of using different birth cohorts
(i.e., different values that may affect SD scores, different academic
levels that may affect AC). Further research is therefore needed to
verify whether the effects reported above can be reproduced in a
longitudinal study.
Acknowledgments
This research was supported by a grant from the Spanish
Ministry of Economy and Competitivity (PSI2011-22683).
References
Anguiano-Carrasco, C., & Vigil-Colet, A. (2011). Assessing indirect
aggression in aggressors and targets: Spanish adaptation of Indirect
Aggression Scales. Psicothema, 23, 146-152.
Anguiano-Carrasco, C., Vigil-Colet, A., & Ferrando, P.J. (2013). Controlling
social desirability may attenuate faking effects: A study with aggression
measures. Psicothema, 25, 164-170.
214
Archer, J. (2004). Which attitudinal measures predict trait aggression?
Personality and Individual Differences, 36, 47-60.
Becker, G. (2007). The Buss-Perry Aggression Questionnaire: Some
unfinished business. Journal of Research in Personality, 41, 434-452.
Biaggio, M. K. (1980). Assessment of anger arousal. Journal of Personality
Assessment, 44, 289-298.
The effects of ageing on self-reported aggression measures are partly explained by response bias
Caspi, A., Roberts, B.W., & Shiner, R.L. (2005). Personality development:
Stability and change. Annual Review of Psychology, 56, 453-484.
Dijkstra, W., Smit, J.H., & Comijs, H.C. (2001). Using social desirability
scales in research among the elderly. Quality & Quantity 35, 107-115.
Ferrando, P.J., Lorenzo-Seva, U., & Chico, E.(2009). A general factoranalytic procedure for assessing response bias in questionnaire
measures. Structural Equation Modeling, 16, 364-381.
Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Mini-Mental State:
A practical method for grading the cognitive state of patients for the
clinician. Journal of Psychiatric Research, 12, 189-198.
Gerevich, J., Bácskai, E., & Czobor, P. (2007). The generalizability of
the Buss-Perry Aggression Questionnaire. International Journal of
Methods in Psychiatric Research, 16, 124-136.
Harris, J.A. (1997). A further evaluation of the Aggression Questionnaire:
Issues of validity and reliability. Behaviour Research and Therapy, 35,
1047-1053.
Krahé, B., & Fenske, I. (2002). Predicting aggressive driving behavior: The
role of macho personality, age, and power of car. Aggressive Behavior,
28, 21-29.
Leite, W.L., & Cooper, L.A. (2010). Detecting social desirability bias using
factor mixture models. Multivariate Behavioral Research, 45, 271-293.
Li, A., & Bagger, J. (2006). Using the BIDR to distinguish the effects of
impression management and self-deception on the criterion validity
of personality measures: A meta-analysis. International Journal of
Selection & Assessment, 14, 131-141.
Lobo, A., Saz, P., Marcos, G., Día, J.L., De la Cámara, C., Ventura, T., ...,
Aznar, S. (1999). Re-validation of the Mini-Examen Cognoscitivo (first
Spanish version of the Mini-Mental Status Examination) in the elderly
people. Medicina Clínica, 112, 767-774.
Lorenzo-Seva, U., & Ferrando, P.J. (2009). Acquiescent responding
in partially balanced multidimensional scales. British Journal of
Mathematical and Statistical Psychology, 62, 319-326.
Marsh, H.W., Nagengast, B., & Morin, A.J. (2013). Measurement
invariance of big-five factors over the life span: ESEM tests of gender,
age, plasticity, maturity, and la dolce vita effects. Developmental
Psychology, 49, 1194-1218.
McCrae, R.R., & Costa, P.T. (1983). Social desirability scales: More
substance than style. Journal of Consulting and Clinical Psychology,
51, 882-888.
Morales-Vives, F., & Vigil-Colet, A. (2010). Are there sex differences
in physical aggression in the elderly? Personality and Individual
Differences, 49, 659-662.
Morales-Vives, F., & Vigil-Colet, A. (2012). Are old people so gentle?
Functional and dysfunctional impulsivity in the elderly. International
Psychogeriatrics, 24, 465-471.
Morren, M., & Meesters, C. (2002). Validation of the Dutch version of
the Aggression Questionnaire in adolescent male offenders. Aggressive
Behavior, 28, 87-96.
O’Connor, D.B., Archer, J., & Wu, F.W. (2001). Measuring aggression:
Self-reports, partner reports, and responses to provoking scenarios.
Aggressive Behavior, 27, 79-101.
Ones, D.S., Viswesvaran, C., & Reiss, A.D. (1996). Role of social
desirability in personality testing for personnel selection: The red
herring. Journal of Applied Psychology, 81, 660-679.
Paulhus, D.L. (1991) Measurement and control of response bias. In
J.P. Robinson, P.R. Shaver & L.S. Wrightsman (Eds.), Measures of
personality and social psychological attitudes (pp. 17-59). San Diego.
Academic Press.
Ray, J.J. (1988). Lie scales and the elderly. Personality and Individual
Differences, 9, 417-418.
Ruiz-Pamies, M., Lorenzo-Seva, U., Morales-Vives, F., Cosi, S., &
Vigil-Colet, A. (2014). I-DAQ: A new test to assess direct and indirect
aggression free of response bias. The Spanish Journal of Psychology,
17, 1-8.
Selby, M.J. (1984). Assessment of violence potential using measures
of anger, hostility, and social desirability. Journal of Personality
Assessment, 48, 531-544.
Soto, C.J., John, O.P., Gosling, S.D., & Potter, J. (2011). Age differences in
personality traits from 10 to 65: Big five domains and facets in a large
cross-sectional sample. Journal of Personality and Social Psychology,
100, 330-348.
Soubelet, A., & Salthouse, T.A. (2011). Influence of social desirability
on age differences in self-reports of mood and personality. Journal of
Personality, 79, 741-762.
Stöber, J. (2001). The Social Desirability Scale-17 (SDS-17) convergent
validity, discriminant validity, and relationship with age. European
Journal of Psychological Assessment, 17, 222-232.
Terracciano, A., McCrae, R.R., Brant, L.J., & Costa, P.T., Jr. (2005).
Hierarchical linear modeling analyses of the NEO-PI-R scale in the
Baltimore longitudinal study of aging. Psychology and Aging, 20, 493506.
Vigil-Colet, A., Ruiz-Pamies, M., Anguiano-Carrasco, C., & Lorenzo-Seva,
U. (2012). The impact of social desirability on psychometric measures
of aggression. Psicothema, 24, 310-315.
Vigil-Colet, A., Morales-Vives, F., Camps, E., Tous, J., & Lorenzo-Seva,
U. (2013). Development and validation of the Overall Personality
Assessment Scales (OPERAS). Psicothema, 25, 100-106.
Vigil-Colet, A., Morales-Vives, F., & Lorenzo-Seva, U. (2013). How social
desirability and acquiescence affect the age-personality relationship.
Psicothema, 25, 342-348.
Walker, S., & Richardson, D.R. (1998). Aggression strategies among older
adults: Delivered but not seen. Aggression and Violent Behavior, 3,
287-294.
Walker, S., Richardson, D.S., & Green, L.R. (2000). Aggression among
older adults: The relationship of interaction networks and gender role to
direct and indirect responses. Aggressive Behavior, 26, 145-154.
215
Descargar